Exploring Elixir for Machine Learning with Savannah Manning and Bruce Tate

In this 27-minute talk, Savannah Manning and Bruce Tate discuss how Elixir can be effectively utilized in machine learning, comparing the process to a climbing expedition. They start by narrating a gripping story of a climbing adventure to draw parallels between the meticulous planning required for climbing and the various stages in machine learning: obtaining, scrubbing, exploring, modeling, and interpreting data. They delve into the specifics of Elixir's machine learning capabilities, focusing on libraries such as Livebook, NX, Axon, and Bumblebee. Bruce Tate explains neural networks and the concept of auto-differentiation for optimization within Elixir. Savannah Manning provides real-world examples of using Bumblebee for various machine learning applications, highlighting the ease and efficiency brought by these tools in Elixir. The talk concludes with a reminder of the importance of community-driven tools and contributions, both in climbing and in the Elixir ecosystem.